Follow applicable law
Do not use RAYA services to facilitate unlawful activity, evade legal obligations, or infringe the rights of others.
15 / ACCEPTABLE USERAYA.COM.AI
A responsible boundary for RAYA services.
This pre-launch policy describes the categories of use RAYA does not intend to support through its models, Token API, customization, or enterprise delivery.
POSITION
Localized capability should expand useful work in Indonesia without enabling abuse, unlawful activity, deception, exploitation, or the removal of human accountability.
RAYA OPERATING PRINCIPLESJAKARTA / INDONESIA
01
Local depth
Language and context are part of the model, not decorative localization.
02
Open progress
Use global open-model innovation to move local capability forward faster.
03
Commercial utility
Build around measurable work for developers and enterprises.
04
Explicit control
Keep data, evaluation, deployment, and responsibility visible.
01 / APPROACH
How it works
Prevent serious harm
Do not use models to meaningfully facilitate violence, exploitation, malicious cyber activity, or dangerous wrongdoing.
Do not deceive
Do not impersonate people, conduct fraud, manipulate users, or present generated content as verified human judgment where that distinction matters.
Keep people responsible
High-impact decisions require appropriate human oversight, validation, escalation, and responsibility.
02 / SCOPE
What it includes
Privacy and identity
No unauthorized collection, exposure, inference, or misuse of personal or confidential information
→Cybersecurity
No malware, credential theft, unauthorized access, destructive activity, or evasion of security controls
→Exploitation
No sexual exploitation, abuse of minors, trafficking, coercion, or targeted facilitation of harm
→High-impact domains
No fully automated consequential decisions without appropriate authority, review, validation, and recourse
→Model integrity
No attempts to bypass technical safeguards, extract protected system information, or interfere with service operation
→Enforcement
RAYA may restrict access, investigate suspected abuse, preserve relevant records, or cooperate with lawful requests
→03 / QUESTIONS
Frequently asked
Is this the final production policy?
It is a pre-launch policy direction. A version tied to the actual service, enforcement process, and legal operator will apply before general availability.
Can RAYA be used in regulated industries?
Potentially, but the organization remains responsible for applicable law, validation, controls, human review, and decisions. Model access alone does not make a workflow compliant.
How should suspected abuse be reported?
Email hello@raya.com.ai with ABUSE or SECURITY in the subject line and provide only the information necessary to investigate safely.